bioRxiv preprint doi: https://doi.org/10.1101/2021.04.14.439731; this version posted April 14, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Jones, Rubin, Dudchenko et al., 2021 – preprint version - bioRxiv

Convergent selection on juvenile hormone signaling is associated with the evolution of in bees

Beryl M. Jones1,2,*, Benjamin E.R. Rubin1,2,*, Olga Dudchenko3,4,*, Karen M. Kapheim5, Eli S. Wyman1,2, Brian G. St. Hilaire3, Weijie Liu1,2, Lance R. Parsons2, S. RaElle Jackson2, Katharine Goodwin2, Shawn M. Davidson2, Callum J. Kingwell6,7, Andrew E. Webb1,2, Mauricio Fernández Otárola8, Melanie Pham3, Arina D. Omer3, David Weisz3, Joshua Schraiber9,10, Fernando Villanea9,11, William T. Wcislo7, Robert J. Paxton12,13, Brendan G. Hunt14, Erez Lieberman Aiden3,4,15,16,ⱡ, Sarah D. Kocher1,2,ⱡ,§

*These authors contributed equally ⱡThese authors contributed equally §Corresponding Author: [email protected]

1Department of Ecology and Evolutionary Biology, Princeton University, USA 2Lewis-Sigler Institute for Integrative Genomics, Princeton University, USA 3The Center for Genome Architecture, Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, USA 4Center for Theoretical Biological Physics, Rice University, USA 5Department of Biology, Utah State University, USA 6Department of Neurobiology and Behavior, Cornell University, USA 7Smithsonian Tropical Research Institute, 0843-03092 City, Republic of Panama 8Research Center for Biodiversity and Tropical Ecology (CIBET), School of Biology, University of Costa Rica, San José, Costa Rica 9Department of Biology, Temple University, USA 10Illumina Artificial Intelligence Laboratory, Illumina Inc, San Diego, CA, USA 11Department of Anthropology, University of Colorado Boulder, USA 12Institute of Biology, Martin-Luther University Halle-Wittenberg, Germany 13German Centre for Integrative Biodiversity Research (iDiv) Halle-Jena-Leipzig, Germany 14Department of Entomology, University of Georgia, USA 15Shanghai Institute for Advanced Immunochemical Studies, ShanghaiTech, Pudong 201210, China 16Faculty of Science, UWA School of Agriculture and Environment, University of Western Australia, Perth WA 6009, Australia

Abstract Life’s most dramatic innovations, from the emergence of self-replicating molecules to highly-integrated societies, often involve increases in biological complexity. Some groups traverse different levels of complexity, providing a framework to identify key factors shaping these evolutionary transitions. Halictid bees span the transition from individual to group reproduction, with repeated gains and losses of eusociality. We generated chromosome-length genome assemblies for 17 species and searched for genes that both experienced positive selection when eusociality arose and relaxed selection when eusociality was secondarily lost. Loci exhibiting these complementary evolutionary signatures are predicted to carry costs outweighed by their importance for traits in eusocial lineages. Strikingly, these loci included two proteins that bind and transport juvenile hormone (JH) – a key regulator of development and reproduction. Though changes in JH abundance are frequently associated with polymorphisms, the mechanisms coupling JH to novel phenotypes are not well understood. Our results suggest novel links between JH and eusociality arose in halictids by altering transport and availability of JH in a tissue-specific manner, including in the brain. Through genomic comparisons of species encompassing both the emergence and breakdown of eusociality, we provide insights into the mechanisms targeted by selection to shape a key evolutionary transition.

Keywords: comparative genomics, convergent evolution, eusociality, behavior, bees, juvenile hormone

Introduction solitary individuals that live and reproduce independently to eusocial Organisms situated at the inflection point of life’s major evolutionary nests where individuals cooperate to reproduce as a group, and even transitions provide a powerful framework to examine the factors shaping polymorphic species that produce both solitary and social nests. This the evolution of these traits1,2. Halictid bees (: ) evolutionary replication enables a comparative approach to identify the offer a unique opportunity to study the evolution of eusociality – social core factors that shape the emergence and breakdown of eusociality and colonies with overlapping generations and a non-reproductive worker provides insights into the most costly aspects of social life. caste. Within the halictid bees, there have been two independent gains3 We searched for signatures of positive selection associated and a dozen losses4 of eusociality. As a result, closely related species with the convergent gains of eusociality as well signatures of relaxed within this group encompass a broad spectrum of social behavior, from selection when eusociality is lost. These complementary patterns

1 bioRxiv preprint doi: https://doi.org/10.1101/2021.04.14.439731; this version posted April 14, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Jones, Rubin, Dudchenko et al., 2021 – preprint version - bioRxiv

indicate that these loci are associated with costs or trade-offs underlying all with chromosome-length scaffolds (Fig. 1; Fig. S1). We selected the maintenance of social traits. We find that some of the targets of species with well-characterized social behaviors that encompass both selection implicated in the origins and elaborations of eusociality, such independent origins and six repeated losses of eusociality3 (Fig. 1a-b). as young, taxonomically restricted genes5 and gene regulatory elements6, Sampling closely-related eusocial and solitary species alongside known also show relaxation of selective pressures when social behavior is lost, non-eusocial outgroups provides a powerful framework to examine the and we uncovered four genes strongly associated with the evolution of molecular mechanisms shaping the evolution of social behavior in this eusociality in halictid bees, including the two primary juvenile hormone group1. binding proteins (JHBPs): apolipoprotein7–9 and hexamerin11010,11. Assemblies ranged in size from 267 to 470 Mb, with estimated These results, combined with our analysis of JH localization in , numbers of chromosomes ranging from 9 to 38 (Table S2). Broadly, we provide new insights into how JH signaling has likely been modified to find that, in contrast to mammalian species (Fig. S2), genomic shape the evolution of eusociality. rearrangements among the bee species occur disproportionately within rather than between chromosomes (see Fig. 1c, Fig. S3). Consequently, A new comparative genomic resource for studying the loci that are on the same chromosome in one bee species also tend to evolution of eusociality occur on the same chromosome in other bee species. This observation is similar to previous findings in dipteran genomes12,13 and may indicate a We generated comparative genomic resources for halictids consisting of broader trend during the evolution of insect chromosomes. 15 de novo genome assemblies and an update of 2 additional assemblies, To increase the quality of genome annotations, we generated

Figure 1. Comparative genomic resources for halictid bees. (a) Halictid bees encompass a wide range of social behaviors, from solitary to eusocial. Solitary species (yellow) nest and reproduce independently. Reproductive queens from eusocial species (blue) initiate colonies with offspring that include non-reproducing workers. Individuals from polymorphic species (green) are capable of both solitary and eusocial reproduction. Non-eusocial outgroups (red) reproduce independently, and include both solitary (Nomia melanderi, Dufourea noveangliae) and communal (Agapostemon virescens) nesting species. (b) Within the , there have been at least two independent gains of eusociality and over a dozen independent losses of eusociality. Colors at branch tips indicate the behavior of each species. Circle sizes are proportional to the total number of orthologs with evidence for positive selection on each terminal branch (abSREL tests in HyPhy16; FDR<0.1). Light blue rectangles denote the two independent gains of social behavior in this family; their width is proportional to the number of orthologs with evidence for positive selection on each origin branch. Proportions of complete, fragmented, and missing BUSCO orthologs are shown for each genome. (c) Genomic data aligned to the outgroup Nomia melanderi. The 14 N. melanderi chromosomes are represented as a circular ideogram with consecutive chromosomes shown in alternating dark and light gray. The area inside the ideogram comprises 18 color-coded circular tracks, each corresponding to one L. calceatum chromosome, and the y-axis of these circular tracks represents the frequency of regions that align to the corresponding region in the N. melanderi genome. Usually, most of the alignments fall into a single “wedge”, indicating that each L. calceatum chromosome corresponds to just one N. melanderi chromosome. The pattern is typical for the bees we studied (see Fig. S3). Outer blue line plot indicates the number of branches in the phylogeny where positive selection was detected at each gene (abSREL, p<0.05). The outermost plot highlights genes that show convergent or complementary patterns of selection. Genes experiencing positive selection on both origin branches are labeled in blue, those experiencing an intensification of selection in extant eusocial lineages and complementary relaxation of selection in secondarily solitary species (HyPhy RELAX, FDR<0.1) are in purple, and genes with a complementary pattern of both positive selection on the origin branches and convergent relaxation of selection associated the loss of eusociality are green (note that not all genes in these categories are annotated in N. melanderi and are therefore not labeled). See methods section for more details.

2 bioRxiv preprint doi: https://doi.org/10.1101/2021.04.14.439731; this version posted April 14, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Jones, Rubin, Dudchenko et al., 2021 – preprint version - bioRxiv

repeated gains3. These reversals provide a powerful lens to identify key genomic factors needed for the maintenance of social living because organisms are expected to minimize investment in traits when social behaviors are lost or unexpressed. This results in the reduction or removal of selective pressures previously maintaining these costly but essential traits22,23. Thus, we predicted that genes associated with trade-offs or costly roles in maintaining eusocial societies should show relaxation of constraint in species that have secondarily reverted to solitary nesting. Consistent with this hypothesis, we found 443 genes showing evidence of convergent, relaxed selection on the six branches representing independent losses of eusociality (HyPhy RELAX20, FDR<0.1; Table S3). These genes are enriched for chromosome condensation (GO:0030261, Fisher’s Exact, q=0.067, enrichment=4.09), indicating that they may play a role in chromosome accessibility and gene regulation24. They are also enriched for vacuolar transport (GO:0007034, Fisher’s Exact, q=0.067, enrichment=3.11; Table S4). To determine whether or not this pattern is unique to the loss of eusociality, we ran the same tests for relaxed selection using extant eusocial lineages as the focal branches. We found 305 genes with Figure 2. Signatures of selection associated with the gains and signatures of relaxation in eusocial species (HyPhy RELAX20, FDR<0.1; losses of eusociality in halictids were identified by testing all orthologs Table S3) enriched for four GO terms related to metabolism (Table S4). for evidence that dN/dS >1 at a proportion of sites on focal branches (Gains 1 and 2 denoted with gray squares 1-2 on the inset; abSREL16 This is a significantly lower proportion of genes experiencing relaxed tests in HyPhy, FDR<0.05). There were 309 orthologs with evidence of selection in eusocial species compared to those experiencing relaxed positive selection in the Augochlorini (Gain 1) and 62 in Halictini s.s. (Gain selection among solitary species (Fisher’s exact test p = 2.42x10-7, odds- 2). Nine of these loci overlapped both origins, which is significantly more ratio = 1.48), suggesting that the loss of eusociality is more often than predicted by chance (enrichment ratio=3.97, Fisher’s exact p=0.004). associated with a release of constraint compared with eusocial Branches representing losses of eusociality (grey square, label 3; yellow 20 maintenance or elaboration. circles) were tested for evidence of relaxed selection (HyPhy RELAX ); Convergent selection. Finally, by comparing genes associated 443 orthologs passed the significance threshold (FDR<0.1), and 4 with the emergence of eusociality to those associated with its loss, we orthologs overlapped all 3 tests (purple), which is substantially more than predicted by chance (Multi-set Exact Test21; fold-enrichment=8.85, have the unique ability to identify some of the most consequential p=0.001). molecular mechanisms associated with social evolution. If a shared set of genes is associated with the emergence and maintenance of social tissue-specific transcriptomes for 11 species, and we also sequenced and behavior in this group, then we would expect to find genes experiencing characterized 1269 microRNAs (miRs) expressed in the brains of 15 both positive selection when eusociality emerges and relaxed selection species (Table S12). The number of annotated genes ranged from 11,060 when social behavior is lost. Indeed, we find four genes that meet these 14 to 14,982, and BUSCO analyses estimated the completeness of our criteria: OG_11519, a gene with no known homologs outside of annotations to range from 93.4 to 99.1% (Fig. 1b; Table S2). Whole- Hymenoptera, Protein interacting with APP tail-1 (PAT1), 15 genome alignments were generated with progressive Cactus (Fig. 1c), apolipoprotein (apolpp), and hexamerin110 (hex110; Fig. 2). This which we then used to identify 52,421 conserved, non-exonic elements overlap is significantly more than expected by chance (Multi-set Exact present in 10 or more species. All genomes, annotations, and alignments Test21; fold-enrichment=8.85, p= 0.001). OG_11519 has no identifiable can be viewed and downloaded from the Halictid Geno me Browser protein domains but is conserved throughout the Hymenoptera. PAT1 (https://beenomes.princeton.edu). modulates mRNA transport and metabolism25,26, and apolpp and hex110 are the primary JH binding proteins and collectively bind nearly all the Convergent and complementary signatures of selection are JH in the hemolymph7–11,27. Hex110 is associated with caste associated with the gains and losses of eusociality differentiation in nearly all examined eusocial insects10,28–30, and ApoLpp has repeatedly been shown to be a major JH binding protein in the Origins. First, to identify convergent signatures of positive 8,9,27 selection associated with transitions from non-eusocial to eusocial life hemolymph . We also find 34 genes that show intensification of selection on extant eusocial lineages and relaxation in secondarily histories, we looked for positive selection on the branches that represent solitary species (HyPhy RELAX, FDR<0.1 for both tests). These genes each origin of eusociality: Augochlorini and Halictini sensu stricto (Fig. likely represent trade-offs associated with the maintenance or elaboration 2). We found 309 genes on the Augochlorini origin branch and 62 genes of eusociality, and they are enriched for regulation of SNARE complex on the Halictini origin branch with evidence of positive selection (HyPhy assembly (GO:0035542, Fisher’s Exact, q=0.074, enrichment=80.91; abSREL16, FDR<0.1) (Table S3). On the Augochlorini branch, genes Table S4), which is a key component of synaptic transmission that has with signatures of positive selection were enriched for cell adhesion also been implicated with variation in social behavior in L. albipes2 and (GO:0007155, Fisher’s Exact, q=2.01e-4; enrichment=2.92; Table S4). 31 There was no detectable gene ontology (GO) enrichment among wasps . positively-selected genes on the Halictini branch (Fisher’s Exact, q>0.1). Nine genes showed signatures of positive selection on both branches Young genes and gene regulatory expansions are associated (Fisher’s Exact, p= 0.004, enrichment=3.97), including genes known to with trade-offs underlying the maintenance of eusociality 17 regulate lifespan and starvation resistance (methuselah-like 10 ) and Previous studies of eusociality have suggested that, similar to their 18 neuronal functions (shopper ), among others (Table S3). importance in the evolution of other novel traits32, younger or Losses. A unique attribute of halictid bees is that there have taxonomically-restricted genes (TRGs) may play key roles in the been a number of independent losses of eusociality19 in addition to

3 bioRxiv preprint doi: https://doi.org/10.1101/2021.04.14.439731; this version posted April 14, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Jones, Rubin, Dudchenko et al., 2021 – preprint version - bioRxiv evolution of eusocial behavior33. To test this hypothesis, we examined the relationship between gene age and selection associated with eusocial origins, maintenance, and reversions to solitary life histories (Fig. S4). We found a greater proportion of young genes compared with old genes experience relaxed selection when eusociality is subsequently lost (Fig. 3A; Pearson’s r=0.869, p=0.002). This relationship does not hold for orthologs showing evidence of relaxed selection on eusocial branches, nor are younger genes more likely than older genes to experience intensified selection associated with either the origins or maintenance of eusociality (Fig. S4) (Table S5). These results suggest that the reversion to solitary life histories is uniquely associated with relaxed selection on young genes, implicating TRGs in traits that are subject to reduced functional constraint when eusociality is lost. Thus, our results partially support other studies in complex eusocial hymenopterans, including honey bees, ants, and wasps, in which young TRGs have been linked to Figure 3. Young genes and increased transcription factor motif the evolution of the non-reproductive workers5,34–37, and they further presence associated with the maintenance of eusociality. (a) Younger suggest that younger genes are associated with costs or trade-offs linked genes are more likely to show signatures of relaxed selection when social to eusociality. behavior is lost. Circles show the proportion of the genes in each age class Gene regulatory changes have also been implicated in eusocial that show evidence of relaxed selection in solitary (yellow) or social (blue) evolution37, including the expansion of transcription factor (TF) motifs lineages, from the oldest Bilaterian group (Age=1) to the youngest, halictid- in the genomes of eusocial species compared with distantly-related specific taxonomically restricted genes (Age=9; see supplement for 6 complete list). For solitary lineages, we find a significant correlation solitary species . To assess the degree to which changes in gene between gene age and the proportion of those genes with evidence of regulation may facilitate the evolution of social behavior in halictids, we relaxed selection (Pearson correlation, r=0.869, p=0.002). This characterized TF motifs in putative promoter regions in each halictid relationship does not hold for the social taxa, where genes experiencing genome. For each species, we defined these regions as 5kb upstream and relaxed selection in these lineages are found at low frequency in all gene 2kb downstream of the transcription start site for each gene6 and age classes (Pearson correlation, r=-0.0447, p=0.91). Shading represents calculated a score for a given TF in each region that reflects the number 95% confidence intervals. (b) STUBB scores38 were calculated for 223 and/or strength of predicted binding sites38. Drosophila transcription factor binding motifs in each halictid genome, and If social species have a greater capacity for gene regulation each motif was tested for overrepresentation in solitary or social genomes; 94 motifs were enriched in social genomes compared to only 29 enriched compared to lineages that have reverted to solitary nesting, then we in solitary genomes. “**” indicates p<0.01. (c) Permutation tests reveal that would expect to find more motifs with scores (reflecting both strength this approximately 3-fold enrichment in (b) is unlikely to occur by chance and number of binding sites) that are higher in social taxa compared to (empirical p=<0.01). secondarily solitary taxa. In support of this hypothesis, we find a greater than 3-fold enrichment of TF motifs that are positively correlated with GO terms (Permutation tests, p<0.05; Table S7; Fig. S6), potentially social lineages compared to secondarily solitary lineages after providing molecular support to hypotheses that the evolution of phylogenetic correction (Fig. 3b), and permutation tests indicate that this eusociality involves altered investments in cognitive function40,41. difference is highly significant (Fig. 3c). Five of these socially-biased motifs were previously identified as associated with eusocial evolution in 6 Convergent signatures of selection on Juvenile Hormone bees , including the motifs for lola, hairy, CrebA, CG5180, and the Binding Proteins (JHBPs) met/tai complex (Table S6). Interestingly, genes showing relaxed selection associated with We identified a small but robust set of genes with complementary the loss of eusociality are enriched for chromosome condensation (Table signatures of selection linked to the repeated gains and losses of S4), suggesting there may be trade-offs or costs associated with this eusociality (Fig. 2). These complementary patterns highlight the increased regulatory potential. To determine whether or not we see a importance of these genes in eusocial lineages: convergent signatures of signature of these trade-offs in non-coding sequences, we used relaxed selection when eusociality is lost suggest they are also associated phastCons39 to identify conserved non-exonic elements (CNEEs) in each with costly trade-offs23. Strikingly, 2 of these 4 genes (apolpp and species. We identified CNEEs that were either slow- or fast- evolving in hex110) are the primary binding proteins for juvenile hormone (JH), an eusocial and solitary lineages. Using the logic outlined above, we would -specific hormone that regulates many aspects of insect life predict that CNEEs essential to the maintenance of social behavior should history including development, reproduction, diapause and be more highly conserved and evolving more slowly in extant eusocial polyphenism42,43. Together, apolpp and hex110 are thought to bind nearly taxa vs. secondarily solitary taxa. After phylogenetic correction, we find all JH in insect hemolymph27. 1,876 CNEEs that meet these criteria, with faster rates of evolution in For both proteins, we find region-specific, faster rates of secondarily solitary species (Fig. S5). This is significantly more than evolution on eusocial branches compared to non-eusocial outgroups. To expected by chance (Binomial test, p=5.27e-9); these regions are proximal identify sites in these genes with evidence of positive selection on the two to genes enriched for several GO terms, including serotonin receptor branches representing independent origins of eusociality, we used a signaling (Permutation tests, p=0.007) and adherens junction mixed-effects maximum likelihood approach (MEME44). For both maintenance (Permutation tests, p=0.009; see Table S7 for complete list). apolpp and hex110, we find evidence for positive selection on sites We also find 1,255 CNEEs that are more highly conserved in secondarily present in functional regions of the protein (Fig. S7), including the solitary lineages (fast in eusocial; slower in solitary). This number is receptor binding domain and predicted binding pocket for ApoLpp as smaller than expected by chance (Binomial test, p<1e-10), but these well as in all three Hemocyanin domains of Hex110 (associated with regions are proximal to genes enriched for cytoskeletal organization and storage functions in these proteins45) and its predicted binding pocket. processes related to synapse and neurotransmitter functions, among other Taken together, these results suggest that positive selection shaped JHBP

4 bioRxiv preprint doi: https://doi.org/10.1101/2021.04.14.439731; this version posted April 14, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Jones, Rubin, Dudchenko et al., 2021 – preprint version - bioRxiv function as eusociality emerged in two independent lineages of these bees, and that some of these changes may be associated with costs when eusociality is lost.

A molecular model for evolutionary novelty in JH signaling JH is essential to reproductive maturation in solitary insects, but this signaling system has also been frequently co-opted during major life history transitions43,46–48, including eusociality42. In relatively small eusocial societies like halictids49,50, paper wasps51, and bumblebees52,53, JH has maintained its association with reproductive maturation, but has also gained a new role in mediating aggression and dominance. In the much more elaborate eusocial colonies of honey bees, further modifications to JH signaling have also resulted in a secondary decoupling of JH in workers independent of its ancestral, reproductive role42. Although the Figure 4. Modifications to JH transport and sequestration could couple behavior with reproduction in association between JH titers and the origins of eusociality. (A) Peak areas of JH III in the brain and hemolymph of four halictid species. In all division of labor is well established in samples, we find detectable levels of JH III in both tissues. n=3 for each tissue of A. pura, A. aurata, and H. ligatus, and n=2 for each tissue from L. versatum. Boxes show median and 25th and 75th percentiles. Whiskers these taxa, we still do not understand show minimum and maximum values. (B) Tissue samples collected from adult bumblebees confirm there is JH which components of the JH signaling III in the hemolymph, brain, and ovaries (Fig. S8, S10). Fresh dissection buffer (PBS) was used as a negative pathways have been targeted or control. (C) Stable isotope tracing reveals that labelled JH III (JH III-D3) applied onto the abdomens of modified by natural selection to bumblebees is transported to the brain and detectable by LC-MS 24 hours later. Bees treated with acetone evolutionarily link JH and social were used as controls. n=3 bee tissues per experimental condition. (D) Hypothetical molecular model. JH is traits. synthesized in the corpora allata (CA) and exerts downstream effects on reproduction and behavior in social 7–9 Our discovery that JHBPs insects. We find convergent positive selection on the two primary JH binding proteins , apolpp and hex110, experience complementary selective when social behavior is gained in halictid bees, and we also find evidence for relaxation of selection on these genes when social behavior is subsequently lost. Hex110 is a storage protein that appears to sequester its pressures when eusociality is gained cargo rather than deliver it to tissues10,28. In contrast, ApoLpp delivers cargo to target tissues, including the or lost leads us to propose a model for brain59,60. Together, these results suggest that these two genes could modulate JH availability in a tissue- how this novel association may arise. specific manner52 – a plausible mechanism by which social behavioral traits (like dominance) can become Specifically, alteration of JHBPs may coupled to ancestrally-reproductive pathways when eusociality arises. Created with BioRender.com. lead to changes in JH availability in a 52 non-reproducing individuals compared to reproductives, bolstering tissue-specific manner , facilitating the decoupling of JH from its 29,30,34,57,58 54 support for the role of this protein in JH sequestration . On the ancestral role in reproduction (Fig. 4). Through modification of binding other hand, ApoLpp (a.k.a. lipophorin, apolipoprotein, or lipoprotein) is affinity and/or altered cellular uptake of JH, evolutionary changes to known to transport its cargo to target tissues via receptor-mediated apolpp and hex110 could modify the overall and tissue-specific levels of endocytosis59 and to cross the blood-brain barrier60. Given that ApoLpp JH55, with differential responses to JH leading to polyphenisms among 54 has repeatedly been shown to be a major JH binding protein in the individuals . Our model is consistent with theory suggesting that hemolymph8,9, it is likely that ApoLpp also plays a role in JH transport conditional expression of a trait (i.e. eusociality) leads to independent 56 and delivery to tissues, including the brain. selection pressures and evolutionary divergence ; evolution of JHBPs Surprisingly, JH has not previously been characterized in the may be one example of such divergence associated with conditional insect brain, though another major insect hormone, ecdysone, was expression of social behavior. recently shown to be actively transported into the brain of There is a growing body of evidence that Hex110 modulates Drosophila61,62. We used liquid chromatography coupled mass JH availability via sequestration. Hex110 is associated with caste 10,28–30 spectrometry (LC-MS) to measure JH titers in several bee species and differentiation in nearly all eusocial insects examined . In a termite, found appreciable levels of JH III in all tissues across species, including Reticulitermes flavipes, RNAi knockdown of hex110 leads to an the brain (Fig. 4a; Fig. S8). In addition, topical application of deuterated increased proportion of reproductives within the colony, suggesting that JH III (JH III-D3) to the abdomens of Bombus impatiens led to Hex110 sequesters JH to prevent it from interacting with downstream 28 incorporation of JH III-D3 into the brain, with an average of 9.5% of total targets . Further, gene expression studies in wasps, ants, and honey bees brain JH being of the deuterated form after 24 hours (Fig. 4b). Our results all reveal that abdominal gene expression levels of hex110 are higher in

5 bioRxiv preprint doi: https://doi.org/10.1101/2021.04.14.439731; this version posted April 14, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Jones, Rubin, Dudchenko et al., 2021 – preprint version - bioRxiv reveal that JH is indeed present in insect brains and can cross the blood- Detig, Nathan Clark, Adam Siepel, and Xander Xue provided discussion brain barrier. and advice on data analysis. We also thank Wenfei Tong for the bee Following transport to relevant tissues, JH binds to the JH drawings and Michael Sheehan for assistance with RNA library receptor, Met, to initiate downstream effects63. Met is a bHLH-PAS preparation. This work was supported by NSF-DEB1754476 awarded to transcription factor that has been well-characterized in flies63,64, SDK and BGH, NIH 1DP2GM137424-01 to SDK, USDA NIFA mosquitoes65, Tribolium flour beetles66, and silk moths67. It is coactivated postdoctoral fellowship 2018-67012-28085 to BERR, DFG PA632/9 to by Taiman (Tai) as the Met/Tai complex, which initiates downstream RJP, a Smithsonian Global Genome Initiative award GGI-Peer-2016-100 transcriptional responses. We identified met/tai as one of the TF motifs to WTW and CK, a Smithsonian Institution Competitive Grants Program correlated with in halictids (Fig. 3b), and we find evidence for for Biogenomics (WTW, KMK, BMJ), a Smithsonian Tropical Research positive selection on tai on the Augochlorini origin branch, suggesting Institute fellowship (to CK), and a gift from Jennifer and Greg Johnson that modifications to JH response-elements have also helped to fine-tune to WTW. MFO was supported by Vicerrectoría de Investigación, UCR, JH signaling in this group of bees. project B7287. ELA was supported by an NSF Physics Frontiers Center This model provides a testable hypothesis for how JH signaling Award (PHY1427654), the Welch Foundation (Q-1866), a USDA could be co-opted during evolutionary transitions to and from eusociality. Agriculture and Food Research Initiative Grant (2017-05741), and an Our findings suggest that the JHBPs, hex110 and apolpp, may function NIH Encyclopedia of DNA Elements Mapping Center Award together to alter the availability and uptake of JH by the brain, ovaries, (UM1HG009375). and other tissues, and that modifications to these proteins could generate novel relationships between behavioral and reproductive traits – a key Permit Information feature of the origins and evolution of eusociality54. Sample permits for collection and import were obtained by SDK, ESW, and MFO (R-055-2017-OT-CONAGEBIO), SDK (P526P-15-04026, and Conclusions RJP (Belfast City Council, Parks & Leisure Dept).

Sweat bees repeatedly traversed an evolutionary inflection point between Data Availability a solitary lifestyle and a caste-based eusocial one with multiple gains and losses of this trait. We developed powerful new comparative genomic All sequence data is available in NCBI’s SRA database under bioprojects resources for this group and identified complementary signatures of PRJNA613468, PRJNA629833, and PRJNA718331. Accession numbers convergent selection associated with the emergence and breakdown of are given in Table S1. Hi-C data is available on www.dnazoo.org. Code eusociality. Factors associated with the origins or maintenance of and additional information are available at eusociality are also associated with its loss, indicating that there may be https://github.com/kocherlab/HalictidCompGen. All genomes, trade-offs, constraints, or costs associated with these genomic changes. annotations, and alignments can be viewed and downloaded from the Strikingly, we find that the functional domains of juvenile hormone Halictid Genome Browser (beenomes.princeton.edu). binding proteins have been modified by selection as eusociality has been gained and lost in halictids, and that these modifications are likely to References affect the transport and availability of JH. Coupled with our finding that 1. Kocher, S. D. & Paxton, R. J. Comparative methods offer powerful insights JH is present in the insect brain, our results help to explain how the into social evolution in bees. Apidologie 45, 289–305 (2014). evolution of a major transition is driven by novel linkages between social 2. Kocher, S. D. et al. The genetic basis of a social polymorphism in halictid behaviors and reproduction that have repeatedly evolved alongside bees. Nat. Commun. 9, 4338 (2018). eusociality in bees. 3. Gibbs, J., Brady, S. G., Kanda, K. & Danforth, B. N. Phylogeny of halictine bees supports a shared origin of eusociality for Halictus and Lasioglossum Author Contributions (Apoidea: Anthophila: Halictidae). Mol. Phylogenet. Evol. 65, 926–939 (2012). BERR, BMJ, and SDK designed the study. BERR, BMJ, and SDK drafted the initial manuscript. Sample collections were performed by 4. Danforth, B. N., Conway, L. & Ji, S. Phylogeny of eusocial Lasioglossum reveals multiple losses of eusociality within a primitively eusocial clade of BERR, ESW, BMJ, SDK, MFO, KMK, and RJP. Genomic and bees (Hymenoptera: Halictidae). Syst. Biol. 52, 23–36 (2003). transcriptomic libraries were generated by BERR and CK. Hi-C sample 5. Johnson, B. R. & Tsutsui, N. D. Taxonomically restricted genes are processing, library preparation, and assembly: OD, BGS, MP, ADO, and associated with the evolution of sociality in the honey bee. BMC Genomics DW. Genome annotation and alignment was performed by BERR and 12, 164 (2011). LP. Genomic data was analyzed by BERR, BMJ, SDK, OD, KMK, BGH, 6. Kapheim, K. M. et al. Social evolution. Genomic signatures of evolutionary JS, and FV. Database generation and curation was performed by AEW. transitions from solitary to group living. Science 348, 1139–1143 (2015). Laboratory experiments were performed by WL and ESW; LC-MS 7. Engelmann, F. & Mala, J. The interactions between juvenile hormone (JH), experiments and analyses conducted by SRJ, KEG, and SMD. ELA lipophorin, vitellogenin, and JH esterases in two cockroach species. Insect supervised Hi-C data generation and analysis. SDK supervised the Biochem. Mol. Biol. 30, 793–803 (2000). project. All authors contributed to the work presented in this manuscript, 8. de Kort, C. A. D. & Koopmanschap, A. B. Molecular characteristics of discussed the results and implications of this study, and commented on lipophorin, the juvenile hormone-binding protein in the hemolymph of the the manuscript. Colorado potato beetle. Arch. Insect Biochem. Physiol. 5, 255–269 (1987).

9. Sevala, V. L., Bachmann, J. A. S. & Schal, C. Lipophorin: A hemolymph Acknowledgments juvenile hormone binding protein in the german cockroach, Blattella We would like to thank our many colleagues that contributed samples germanica. Insect Biochem. Mol. Biol. 27, 663–670 (1997). and field support to this dataset, including: Jason Gibbs, Sam Droege, 10. Martins, J. R., Nunes, F. M. F., Cristino, A. S., Simões, Z. L. P. & Bitondi, Raphael Jeanson, Joan Milam, Jakub Straka, Marion Podolak, James M. M. G. The four hexamerin genes in the honey bee: structure, molecular Cane, and Mallory Hagadorn. We also thank Miriam Richards, Cecile evolution and function deduced from expression patterns in queens, workers Plateaux-Quenu, Jason Gibbs, and Laurence Packer for discussion and and drones. BMC Mol. Biol. 11, 23 (2010). insights on halictid life history and behavior. Tim Sackton, Russ Corbett-

6 bioRxiv preprint doi: https://doi.org/10.1101/2021.04.14.439731; this version posted April 14, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Jones, Rubin, Dudchenko et al., 2021 – preprint version - bioRxiv

11. Ismail, S. M. & Gillott, C. Identification, characterization, and 33. Johnson, B. R. Taxonomically restricted genes are fundamental to biology developmental profile of a high molecular weight, juvenile hormone-binding and evolution. Front. Genet. 9, 407 (2018). protein in the hemolymph of the migratory grasshopper, Melanoplus 34. Warner, M. R., Qiu, L., Holmes, M. J., Mikheyev, A. S. & Linksvayer, T. A. sanguinipes. Arch. Insect Biochem. Physiol. 29, 415–430 (1995). Convergent eusocial evolution is based on a shared reproductive groundplan 12. Nene, V. et al. Genome sequence of Aedes aegypti, a major arbovirus vector. plus lineage-specific plastic genes. Nat. Commun. 10, 2651 (2019). Science 316, 1718–1723 (2007). 35. Harpur, B. A. et al. Population genomics of the honey bee reveals strong 13. Dudchenko, O. et al. De novo assembly of the Aedes aegypti genome using signatures of positive selection on worker traits. Proc. Natl. Acad. Sci. U. S. Hi-C yields chromosome-length scaffolds. Science 356, 92–95 (2017). A. 111, 2614–2619 (2014). 14. Simão, F. A., Waterhouse, R. M., Ioannidis, P., Kriventseva, E. V. & 36. Ferreira, P. G. et al. Transcriptome analyses of primitively eusocial wasps Zdobnov, E. M. BUSCO: assessing genome assembly and annotation reveal novel insights into the evolution of sociality and the origin of completeness with single-copy orthologs. Bioinformatics 31, 3210–3212 alternative phenotypes. Genome Biol. 14, R20 (2013). (2015). 37. Simola, D. F. et al. Social insect genomes exhibit dramatic evolution in gene 15. Armstrong, J. et al. Progressive alignment with Cactus: a multiple-genome composition and regulation while preserving regulatory features linked to aligner for the thousand-genome era. Cold Spring Harbor Laboratory sociality. Genome Res. 23, 1235–1247 (2013). 730531 (2019) doi:10.1101/730531. 38. Sinha, S., Liang, Y. & Siggia, E. Stubb: a program for discovery and analysis 16. Smith, M. D. et al. Less is more: an adaptive branch-site random effects of cis-regulatory modules. Nucleic Acids Res. 34, W555–9 (2006). model for efficient detection of episodic diversifying selection. Mol. Biol. 39. Siepel, A. et al. Evolutionarily conserved elements in vertebrate, insect, Evol. 32, 1342–1353 (2015). worm, and yeast genomes. Genome Res. 15, 1034–1050 (2005). 17. Sung, E. J. et al. Cytokine signaling through Drosophila Mthl10 ties lifespan 40. Dunbar, R. I. M. The social brain hypothesis. Evol. Anthropol. 6, 178–190 to environmental stress. Proc. Natl. Acad. Sci. U. S. A. 114, 13786–13791 (1998). (2017). 41. O’Donnell, S. et al. Distributed cognition and social brains: reductions in 18. Otto, N. et al. The sulfite oxidase Shopper controls neuronal activity by mushroom body investment accompanied the origins of sociality in wasps regulating glutamate homeostasis in Drosophila ensheathing glia. Nat. (Hymenoptera: Vespidae). Proceedings of the Royal Society B: Biological Commun. 9, 3514 (2018). Sciences 282, 20150791 (2015). 19. Wcislo, W. T. & Danforth, B. N. Secondarily solitary: the evolutionary loss 42. Hartfelder, K. Insect juvenile hormone: from ‘status quo’ to high society. of social behavior. Trends Ecol. Evol. 12, 468–474 (1997). Braz. J. Med. Biol. Res. 33, 157–177 (2000). 20. Wertheim, J. O., Murrell, B., Smith, M. D., Kosakovsky Pond, S. L. & 43. Nijhout, H. F. & Wheeler, D. E. Juvenile hormone and the physiological Scheffler, K. RELAX: detecting relaxed selection in a phylogenetic basis of insect polymorphisms. Q. Rev. Biol. 57, 109–133 (1982). framework. Mol. Biol. Evol. 32, 820–832 (2015). 44. Murrell, B. et al. Detecting individual sites subject to episodic diversifying 21. Wang, M., Zhao, Y. & Zhang, B. Efficient test and visualization of multi-set selection. PLoS Genet. 8, e1002764 (2012). intersections. Sci. Rep. 5, 16923 (2015). 45. Burmester, T. Evolution and function of the insect hexamerins. Eur. J. 22. Wittwer, B. et al. Solitary bees reduce investment in communication Entomol. 96, 213–226 (1999). compared with their social relatives. Proc. Natl. Acad. Sci. U. S. A. 114, 6569–6574 (2017). 46. Corbitt, T. S. & Hardie, J. Juvenile hormone effects on polymorphism in the pea aphid, Acyrthosiphon pisum. Entomol. Exp. Appl. 38, 131–135 (1985). 23. Lahti, D. C. et al. Relaxed selection in the wild. Trends Ecol. Evol. 24, 487– 496 (2009). 47. Zera, A. J. & Denno, R. F. Physiology and ecology of dispersal polymorphism in insects. Annu. Rev. Entomol. 42, 207–230 (1997). 24. Martin, R. M. & Cardoso, M. C. Chromatin condensation modulates access and binding of nuclear proteins. FASEB J. 24, 1066–1072 (2010). 48. Hartfelder, K. & Emlen, D. J. 11 - Endocrine Control of Insect Polyphenism. in Insect Endocrinology (ed. Gilbert, L. I.) 464–522 (Academic Press, 2012). 25. Loiseau, P., Davies, T., Williams, L. S., Mishima, M. & Palacios, I. M. Drosophila PAT1 is required for Kinesin-1 to transport cargo and to 49. Smith, A. R., Kapheim, K. M., Pérez-Ortega, B., Brent, C. S. & Wcislo, W. maximize its motility. Development 137, 2763–2772 (2010). T. Juvenile hormone levels reflect social opportunities in the facultatively eusocial sweat bee genalis (Hymenoptera: Halictidae). Horm. 26. Lobel, J. H., Tibble, R. W. & Gross, J. D. Pat1 activates late steps in mRNA Behav. 63, 1–4 (2013). decay by multiple mechanisms. Proc. Natl. Acad. Sci. U. S. A. 116, 23512– 23517 (2019). 50. Bell, W. J. Factors controlling initiation of vitellogenesis in a primitively social bee, Lasioglossum zephyrum (Hymenoptera: Halictidae). Insectes Soc. 27. Hidayat, P. & Goodman, W. G. Juvenile hormone and hemolymph juvenile 20, 253–260 (1973). hormone binding protein titers and their interaction in the hemolymph of fourth stadium Manduca sexta. Insect Biochem. Mol. Biol. 24, 709–715 51. Tibbetts, E. A. & Izzo, A. S. Endocrine mediated phenotypic plasticity: (1994). condition-dependent effects of juvenile hormone on dominance and fertility of wasp queens. Horm. Behav. 56, 527–531 (2009). 28. Zhou, X., Tarver, M. R. & Scharf, M. E. Hexamerin-based regulation of juvenile hormone-dependent gene expression underlies phenotypic plasticity 52. Shpigler, H. Y. et al. Juvenile hormone regulates brain-reproduction tradeoff in a social insect. Development 134, 601–610 (2007). in bumble bees but not in honey bees. Horm. Behav. 126, 104844 (2020). 29. Hawkings, C., Calkins, T. L., Pietrantonio, P. V. & Tamborindeguy, C. 53. Pandey, A., Motro, U. & Bloch, G. Juvenile hormone interacts with multiple Caste-based differential transcriptional expression of hexamerins in response factors to modulate aggression and dominance in groups of orphan bumble to a juvenile hormone analog in the red imported fire ant (Solenopsis invicta). bee (Bombus terrestris) workers. Horm Behav 117, 104602 (2020). PLoS One 14, e0216800 (2019). 54. West-Eberhard, M. J. Wasp societies as microcosms for the study of 30. Hunt, J. H. et al. A diapause pathway underlies the gyne phenotype in development and evolution. Natural history and evolution of paper wasps Polistes wasps, revealing an evolutionary route to caste-containing insect 290, 317 (1996). societies. Proc. Natl. Acad. Sci. U. S. A. 104, 14020–14025 (2007). 55. Nijhout, H. F. & Reed, M. C. A mathematical model for the regulation of 31. Wyatt, C. D. R. et al. Genetic toolkit for sociality predicts castes across the juvenile hormone titers. J. Insect Physiol. 54, 255–264 (2008). spectrum of social complexity in wasps. Cold Spring Harbor Laboratory 56. West-Eberhard, M. J. Phenotypic plasticity and the origins of diversity. 2020.12.08.407056 (2020) doi:10.1101/2020.12.08.407056. Annu. Rev. Ecol. Syst. 20, 249–278 (1989). 32. Chen, S., Krinsky, B. H. & Long, M. New genes as drivers of phenotypic evolution. Nat. Rev. Genet. 14, 645–660 (2013).

7 bioRxiv preprint doi: https://doi.org/10.1101/2021.04.14.439731; this version posted April 14, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Jones, Rubin, Dudchenko et al., 2021 – preprint version - bioRxiv

57. Colgan, T. J. et al. Polyphenism in social insects: insights from a 63. Charles, J.-P. et al. Ligand-binding properties of a juvenile hormone transcriptome-wide analysis of gene expression in the life stages of the key receptor, Methoprene-tolerant. Proc. Natl. Acad. Sci. U. S. A. 108, 21128– pollinator, Bombus terrestris. BMC Genomics 12, 623 (2011). 21133 (2011). 58. Hoffman, E. A. & Goodisman, M. A. D. Gene expression and the evolution 64. Miura, K., Oda, M., Makita, S. & Chinzei, Y. Characterization of the of phenotypic diversity in social wasps. BMC Biol. 5, 23 (2007). Drosophila Methoprene-tolerant gene product: Juvenile hormone binding 59. Van Hoof, D., Rodenburg, K. W. & van der Horst, D. J. Lipophorin receptor- and ligand-dependent gene regulation. FEBS J. 272, 1169–1178 (2005). mediated lipoprotein endocytosis in insect fat body cells. J. Lipid Res. 44, 65. Li, M., Mead, E. A. & Zhu, J. Heterodimer of two bHLH-PAS proteins 1431–1440 (2003). mediates juvenile hormone-induced gene expression. Proc. Natl. Acad. Sci. 60. Brankatschk, M. & Eaton, S. Lipoprotein particles cross the blood-brain U. S. A. 108, 638–643 (2011). barrier in Drosophila. J. Neurosci. 30, 10441–10447 (2010). 66. Zhang, Z., Xu, J., Sheng, Z., Sui, Y. & Palli, S. R. Steroid receptor co- 61. Okamoto, N. et al. A Membrane transporter is required for steroid hormone activator is required for juvenile hormone signal transduction through a uptake in Drosophila. Dev. Cell 47, 294–305.e7 (2018). bHLH-PAS transcription factor, Methoprene Tolerant. J. Biol. Chem. 286, 8437–8447 (2011). 62. Okamoto, N. & Yamanaka, N. Steroid hormone entry into the brain requires a membrane transporter in Drosophila. Curr. Biol. 30, 359–366.e3 (2020). 67. Kayukawa, T. et al. Transcriptional regulation of juvenile hormone- mediated induction of Krüppel homolog 1, a repressor of insect metamorphosis. Proc. Natl. Acad. Sci. U. S. A. 109, 11729–11734 (2012).

8